Compelling data support a genetic basis for type 2 diabetes (T2DM), and completed genome scans suggest at least 7 likely susceptibility regions. A major resource to map genes for T2DM was assembled in a multicenter, multiethnic study funded by the American Diabetes Association (GENNID) from 1993-2003 which comprises over 6000 individuals including 770 African-American, 1180 Caucasian, and 1190 Hispanic sib pairs. Under 1/3 of this resource has been genotyped. We propose a complete genome scan on all samples using a 0.6 cM single nucleotide polymorphism (SNP) map, to make the genotype, pedigree, and phenotypic data publicly available in both raw and cleaned forms, and to use this resource to test the hypothesis that T2DM will be characterized by multiple interacting loci, some that are ethnic-specific and others that will cross ethnic groups. The 5 Specific Aims will include 1) sample and data preparation for the genome scan; 2) analysis of the genome scan data using multipoint affected sib pair methods applied to both the full data set and ethnic-specific data, along with secondary parametric, ordered subset, and interaction analyses; 3) fine mapping of linkage peaks with dense (100 kb) SNP maps across up to 12 new regions and reanalysis of the fine map data; 4) distribution of initial and fine mapping data to all interested diabetes investigators upon initialCIDR release and at each stage of cleaning or new data analysis; and 5) a positional candidate gene search for likely susceptiblity genes under each linkage peak by choosing tagSNPs from public databases and gene screening where needed. We will test for family based association using the dense SNP map, and for an association in 400-500 unrelated cases drawn from GENNID families and 400- 500 controls. These studies will provide the basis for future proposals to identify the causative mutations in known candidate genes or to use linkage disequilibrium to identify additional T2DM susceptibility loci. The proposed studies will identify the specific genes that increase an individual's risk of getting type 2 diabetes, and will determine whether striking differences in diabetes prevalence across major United States ethnic groups can be explained in part by different frequencies of risk genes. These studies will determine whether different pathways cause type 2 diabetes in different ethnic groups, which in turn may lead to new therapies for type 2 diabetes that may target specific genetic defects or ethnic-specific pathways.